The proposal indicates that the EEG may be the most direct, inexpensive and noninvasive physiological means by which to access the nonlinear neurodynamics of the human brain. Yet, for many years, it has been both neglected and under-utilized. The reason is that EEG waveforms are baffling in their complexity and uncontrolled variability. To most observers, they look like "noise" of the kind made by mechanical calculators or "the roar of crowds at football games". They are widely considered devoid of behaviorally significant information, and therefore unworthy of serious scientific study and analysis. Recent advances in the experimental technologies involving multichannel recording of brain activity in behaving animals have shown that behavioral information does exist in EEGs, not (as already well known) in their waveforms in time, but in the spatial patterns observed with arrays of electrodes. Recent advances in nonlinear dynamics and in the mathematical descriptions of chaos have provided a body of theory with which to predict the salient characteristics of these spatial patterns, the ways they are generated by networks of interacting neurons in the cerebral cortex, and the ways in which they are established and removed by processes of associative learning and habituation. The proposed program would test quantitative predictions from the theory about the space-time patterns of chaotic neurodynamic activities which would be extracted and measured in the EEGs recorded from the paleocortex and neocortex of rabbits. The expected outcome is anticipated to be a theory of the nonlinear neurodynamics of somesthetic, auditory, and visual neocortex in the rabbit, that might be extended subsequently to understanding human neocortical function and its sensory, motor, and cognitive malfunctions.